I'm not sure the 779 as enough pins to be something really complex when the 8049 is used and the 8049 is probably almost a complete game in a microcontroler. a very partial pinout of the 779 is (when working with the void cartridge SD201)
---------------- | U | ----| 1 42 |---- Clock 4MHz ----| 2 41 |---- VCC 5.6V ----| 3 40 |---- ----| 4 39 |---- ----| 5 38 |---- ----| 6 37 |---- ? out video 50hz ----| 7 36 |---- ? out video 50hz ----| 8 35 |---- ? out video 50hz ----| 9 34 |---- ? out video 50hz ----|10 33 |---- out video 15khz ----|11 32 |---- ----|12 31 |---- out video 15khz ----|13 30 |---- out video 15khz ----|14 29 |---- out video 15khz ----|15 28 |---- out video 15khz ----|16 27 |---- out video 15khz ----|17 26 |---- out video 15khz ----|18 25 |---- ----|19 24 |---- ----|20 23 |---- GND ----|21 22 |---- ----------------
But now that I have a logic analyser I can make a bit more advanced search. I do not have the 8049 cartridge but I have a motorola (SD203) that is probably also an almost one chip game.
I still think all the A/V stuff is done by the 779 and the carts only hold the game rom, PeT also seem to agree with this when he was having a look at trying to get the SD-204 dump running. From his investigations he told me: "When it is really a 8041/MCS48 variant then the cpu core is already emulated in MAME. The console schematics might be needed for proper understanding of the hardware for emulation. Started emulation, but it doesn't have something like video memory. Instead it transmits high level commands over the bus and the D779C does some rendering."
It think that putting some logic analyser will tell more. My guess (but only an hypothesis for now) is that all cartridges have an MCU that takes hand controller as an input (the void SD201 cartridge is a switch to bring hand controls to the 779, and probably also another switch to discard internal game). I do not think the system has much video memory, graphics may be generated on the fly by mcu in sync with CRT like atari VCS. I have the console and SD203, as soon as I will reinstall my crapy logic analyser I will do some tests (in the next days I think).
Yes, it would be nice if you could test that and try to find out for sure. SlyDC already made some progress on this, but not sure how far he went he never got back to the last email I sent him about the subject.
I can confirm that all connections from both hand controllers go to the cartridge connector. Also uPD779 does video and sound generation (at least for SD203 and all cartridges based on the same PCB). There is a direct connection between uPD779 and the MCU on the cartridge, there are 6 bits from cartridge MCU to uPD779 (throw an hex buffer 14050). I did not check the direction, A great number of the uPD779 pins are unconnected.
Unfortunately I fried the SD203 MCU while connecting the ground of the logic analyser so I will need to find another cartridge (It's common and still cheap here).
Nice, it is as suspected then. I'm sure it will work like that for all carts. Don't forget that SD-205 uses the paddle so some of those pins might be used for the paddle controller. Too bad you fried the SD-203, there's actually another copy for sale on ebay right now, but I suppose you can get it much cheaper there.
the potentiometer from controller also goes to the cartridge port, in SD205 a HEF4069 hex inverter is probably used as an oscillator that is read by the MCU. I should receive next week another console + SD203 + SD201 + SD208. I now know where I should put the analyser (and my 8chanel will be enough for the task)
received a second HMG today. the good new is that my first SD203 is not dead (almost work in the second console but cannot start because of bad contact for the game start button). The defect is MC14050 or the uPD779 input ports. since the 14050 is an old CMOS I hope it is the main problem (I will buy some new and supports today) SD208 is a 8049 (does not start game for the same problem of game start I think)